Our involvement in the regulation and consequences of ubiquitination began in the course of studies aimed at understanding why double positive (DP) thymocytes are so sensitive to pro-apoptotic stimuli. One clue was that DP apoptosis is blocked by proteasome inhibitors, implicating regulated protein degradation as a sensitizing event. We found that induction of DP apoptosis, regardless of the molecular pathway, resulted in the degradation of XIAP and c-IAP1, proteins of the Inhibitor of APoptosis (IAP) family. Importantly, we identified XIAP and c-IAP1 as ubiquitin protein ligases (E3s), enzyme involved in the addition of Ub to target proteins. This activity was dependent upon a motif called the RING domain. In subsequent studies we made the following findings: - Signaling via Tumor Necrosis Factor (TNF) receptor 2 (TNF-R2), but not TNF-R1, results in ubiquitination and degradation of the signaling intermediate TRAF2 (TRAF2 is required for coupling the TNF-R to JNK activation and, with TRAF5, to NF-kappaB). - TNF-R2 -mediated ubiquitination of TRAF2 is mediated by c-IAP1. - Expression of an "E3-dead" c-IAP1 RING point mutant (a dominant negative) prevented TNF-&amp;#945;-induced TRAF2 degradation and inhibited apoptosis, demonstrating that c-IAP1 can actually be pro-apoptotic, probably by causing the degradation of TRAF2 and, perhaps, other anti-apoptotic molecules. - Stimulation via TNF-R2 results in the translocation of a c-IAP1/TRAF2 complex to the perinuclear ER, where it encounters a ubiquitin conjugating enzyme (E2), which cooperates with c-IAP1 to cause TRAF2 ubiquitination. - We were the first to characterize mice deficient in c-IAP1, and found that it has an obligate role in the ubiquitination and degradation of another member of the IAP family, c-IAP2. We have also found that ASK1, an important upstream enzyme in the MAP kinase signaling cascade, is a target for c-IAP1 in B cells stimulated with TNF. As a result, MAP kinase signaling is terminated in a timely fashion, moderating B cell responses. Another area in which we have studied ubiquitination is in signaling for activation of the important transcription factor NF-kappaB. NF-kappaB is sequestered in the cytoplasm in a complex with IkappaB. Almost all NF-kappaB activation pathways converge on IkappaB kinase (IKK), which phosphorylates IkappaB resulting in I&amp;#954;B K48-linked polyubiquitination, IkappaB degradation by proteasomes, and migration of NF-kappaB to the nucleus. IKK has two enzymatically-active subunits, IKKalpha and IKKbeta, and a regulatory subunit, IKKgamma or NEMO. NEMO is essential for NF-kappaB activation, and NEMO mutations or deficiency have been identified as the cause of incontinentia pigmenti (IP) and hypohidrotic ectodermal dysplasia and immunodeficiency (HED-ID). The mechanism by which proximal cytokine receptor signals result in its NEMO-dependent activation remains largely unknown. Among the best-studied of such signaling pathways is that for TNF-alpha. TNF receptor 1 (TNF-R1) occupancy results in receptor trimerization and the serial recruitment of TNF receptor-associated death domain (TRADD), Fas-associated death domain (FADD), receptor-interacting protein (RIP), TRAF2, and c-IAP1 and c-IAP2. RIP in particular is an essential intermediate for downstream activation of NF-&amp;#954;B. Upon stimulation with TNF-alpha, RIP binds to NEMO, which brings with it the other IKK components. The RIP that associates with TNF-R1 undergoes polyubiquitination, initially K63-linked, in lipid rafts;the K63-linked polyUb is subsequently removed by the de-ubiquitinating domain of A20 and K48-linked polyUb chains are added by the zinc finger region of A20, resulting in RIP degradation. We used NEMO in a yeast two-hybrid screen to look for interacting proteins, and identified 2-3 residue multi-ubiquitin as a binding partner. In depth analysis identified a motif in NEMO that bound K63-linked (but not K48-linked) polyUb. Moreover, NEMO specifically bound polyUb-modified RIP in TNF-signaled cells, which was prevented by mutations in the NEMO binding site. These same mutations have been identified as causing HED-ID in humans. This led us to conclude that one major function of NEMO is to act as a sensor of K63-linked polyUb, which explains why polyUb is necessary for signaling for NF-kappaB activation in the TNF pathway. Optineurin is a protein whose mutation is responsible for a subset of adult-onset primary open angle glaucoma. Optineurin contains a motif highly homologous to the Ub-binding motif in NEMO. In fact, we have found the optineurin binds to K63-linked polyUb much better than NEMO, and that it competes with NEMO for ubiquitinated RIP in TNF-stimulated cells. Acquisition of optineurin inhibits NF-&amp;#954;B activation, and forced knock-down of optineurin greatly enhances NF-&amp;#954;B activation. Given that NF-kappaB greatly enhances excitotoxic neuronal cell death, we have proposed that loss-of-function mutations in optineurin may in fact cause glaucoma due to enhanced retinal neuron cell death. We have also been studying other signaling pathways that result in NF-kappaB activation, in particular IL-1 receptor (IL-1R)/Toll-like receptor and T cell receptor (TCR) signaling. We have found that both types of receptor use a strategy similar to that of TNF. In particular, in each a particular molecule in the signaling complex is polyubiquitinated with K63-linked chains, which results in the recruitment of NEMO and activation of IKK and NF-&amp;#954;B. Disruption of this, either by mutating the binding site in NEMO or the ubiquitination sites in the adaptor molecules, greatly diminished NF-&amp;#954;B activation. These results establish our finding that NEMO recognizes K63-linked polyUb chains as a general and evolutionarily conserved mechanism for activating NF-kappaB in response to extracellular stimuli. Given the central importance of NF-&amp;#954;B in immune and inflammatory responses, this identifies NEMO/polyUb binding as a possible molecular target of intervening in these processes. We have now generated mice in which we "knocked-in" an E3-defective c-IAP2 (it contains a point mutation in the RING domain). In the past year we have have found - accumulation of B cells, especially of the marginal zone phenotype, and IgA hypergammaglobulinemia. - increased gut-associated lymphoid tissue (GALT) and lymphocyte inflitrates in the lung. - B cell hyperproliferation and relative insensitivity to growth factor-withdrawal apoptosis. - spontaneous B cell NF-kappaB activity via the non-canonical pathway. This phenotype is very similar to that of human MALT lymphomas, a large fraction of which arise as a result of a chromosomal translocation that truncates the c-IAP2 RING domain, and provides a new, and physiologic, model for this category of neoplasia. - c-IAP1 E3-defective c-IAP1 mice have been generated as well, and are being characterized. We have also generated optineurin knockin mice that lack the first 129 amino acids. We have found that optineurin binds two kinases, Tank-binding kinase 1 and IKKepsilon, which are involved in the production of Type I interferons, and thus anti-viral responses. We are currently asking what are the consequences of this for the immune response.